The Internet is the infrastructure of the global economy. It routes binary traffic between billions of devices in its massive ecosystem of interconnected computer networks that run the world at average speeds of 7.2 megabits per second (averaging 28.6Mbps if you live in South Korea). Quantification of the average human brain’s processing ability is less well defined, but the capacity of working-memory is estimated to manage only four to seven chunks of information at a time. To compensate for mental bandwidth limitations people turn to technology. For decades we have outsourced our intelligence to machines to enhance productivity and profitability. Everyday applications of machine intelligence are deployed in domestic GPS navigators that feed their algorithms with accident, road work and congestion data to optimise traffic flows; autopilot functions in commercial aviation; pricing models and estimated waiting periods of ride sharing apps; fraud and credit checks in financial services; personalisation and recommendations displayed in social networks and e-commerce stores, etc… The success of these systems delivering for customer needs is contingent on interconnecting public and proprietary datasets powered by the Internet.

At this point corporate readers might pause to question the effectiveness of their data strategies. Many have spent small fortunes collecting, hording, cleaning and protecting information with little commercial gain. For some the hype of “Big Data” has failed to live up to expectations, in part because of data silos, which perpetuate the same perspective just in finer detail, and the unwillingness to share and link data with other providers to take advantage of network effects.  The principle winners in the big data game are aggregators, particularly advertising platforms like Google and Facebook, which gather data about everything and competently knit it together (incidentally, these two tech giants invest heavily into artificial intelligence). Quality data is integral to their business models and should be to other businesses too. To generate value from in‑house company datasets the data must be converted to meaningful information, best achieved by applying artificial intelligence (AI).

Typical use cases of applied AI, relate to predictive analytics, forecasting, identifying patterns and outliers, real-time optimisation, deep personalisation and mining unstructured data. In practice two significant AI revenue drivers are: product development and process efficiency. Companies that innovate tend to have greater longevity and financial gains than those that don’t. Innovations may be incremental, disruptive, reverse or breakthrough. Harvesting the capabilities of AI to understand consumers’ relationships with products and services to satisfy their demands will lead to improved offerings, and or the creation of completely new solutions based on different permutations to meet unmet needs, in faster development cycles. In terms of efficiency, automating rote tasks, like data capture or verification, to machines improves accuracy rates and speeds up performance – with the potential to run 24/7, achieve scale and realise cost savings. Forms of single purpose basic assisted intelligence are readily available and largely contribute to improving performance. More complex applications of artificial intelligence, the type used to help postulate novel concepts or reconfigure existing offerings, require more sophisticated machine learning and natural language processing technology. This level of artificial intelligence makes the case of man with machine, where machines do the heavy lifting and people leverage the processing capabilities of computers to make intelligent decisions from data. For competitive companies artificial intelligence must be an essential part of their business models both commercially and culturally. Data will become more abundant and more connections will be made as the ‘Internet of Things’ materialises, but interconnections without intelligent interpretation will be worthless.

Until now we have largely been training machines to imitate human behaviour, because that’s what we know. Is it the right approach? Humans tend to think linearly, even though we operate in increasingly non-linear environments. To succeed in a world that is being transformed by and into digitally connected networks we need to develop the capacity to think in an interconnected manner. Artificial intelligence is the technology that will equip us to think differently; to think digital.

Dr Amaleya Goneos-Malka – Marketing and Corporate Strategy Management